111111
`
`1111111111111111111111111111111111111111111111111111111111111111111111111111
`US 20040141566Al
`
`(19) United States
`(12) Patent Application Publication
`Kim et al.
`
`(10) Pub. No.: US 2004/0141566 A1
`Jul. 22, 2004
`( 43) Pub. Date:
`
`(54) APPARATUS AND METHOD FOR
`ADAPTIVELY MODULATING SIGNAL BY
`USING LAYERED TIME-SPACE DETECTOR
`USED IN MIMO SYSTEM
`
`(76)
`
`Inventors: Young-Doo Kim, Seoul (KR);
`Inhyoung Kim, Daejon (KR); Heejung
`Yu, Daejon (KR); Jihoon Choi,
`Gyeongsangnam-Do (KR); Taehyun
`Jeon, Seoul (KR); Jae-Young Ahn,
`Daejon (KR); Yong-Hoon Lee, Daejon
`(KR)
`
`Correspondence Address:
`BLAKELY SOKOLOFF TAYLOR & ZAFMAN
`12400 WILSHIRE BOULEVARD, SEVENTH
`FLOOR
`LOS ANGELES, CA 90025 (US)
`
`(21) Appl. No.:
`
`10/737,398
`
`(22) Filed:
`
`Dec. 16, 2003
`
`(30)
`
`Foreign Application Priority Data
`
`Dec. 26, 2002
`
`(KR) ....................................... 2002-84478
`
`Publication Classification
`
`Int. CI? ....................................................... H04L l/02
`(51)
`(52) U.S. Cl. .............................................................. 375/267
`
`(57)
`
`ABSTRACT
`
`An apparatus for adaptively modulating/demodulating sig(cid:173)
`nals in a multi-input multi-output (MIMO) system having a
`layered time-space architecture detector and a method
`thereof is disclosed. The apparatus includes: a bit and power
`allocation information calculator for deciding an equivalent
`channel gain in a reverse order of Vertical-Belllaboratories
`Space Time (V-BLAST) based on MIMO channel informa(cid:173)
`tion feedbacked from a receiver and determining the number
`of bits and transmission power to be transmitted to each
`transmitting antenna by using the equivalent channel gain;
`and adaptive modulation means for modulating signal of
`each layer with corresponding modulation method based on
`the determined number of bits and transmitting power,
`controlling the transmitting power and transmitting the
`adaptively modulated signal through each transmitting
`antenna. The present invention can improve performance
`without increase of implementation complexity and easily
`expand to MIMO-OFDM system by adaptively modulating
`and demodulating signals in reverse order of conventional
`V-BLAST detection method.
`
`12
`
`16
`
`ADAPTIVE
`MODULATION
`UNIT
`
`ADAPTIVE
`DEMODULATION
`UNIT
`
`SONY EX. 1005
`Page 1
`
`

`

`Patent Application Publication
`
`Jul. 22, 2004 Sheet 1 of 8
`
`US 2004/0141566 A1
`
`,...__ ...--0----.
`w
`1---
`::So::t::
`::::>1--(cid:173)
`o<t:
`00
`::::2: w
`0
`
`z
`0
`w I-
`><(
`-_JI-
`1-:::::>-
`o_oz
`C§o:::::>
`<(~ w
`0
`
`<0
`
`I
`lO
`
`N
`I
`lO
`
`z
`I
`lO
`
`.
`
`. .
`
`.
`
`.
`
`.
`C)
`u....
`
`I
`(V)
`
`N
`I
`(V)
`
`<..'J
`a:::Z:
`01--(cid:173)
`u....l---
`o::t::::::2:
`I- (/)
`<t: :z:
`0 <t: a:
`I-
`
`SONY EX. 1005
`Page 2
`
`

`

`'"""
`>
`0'1
`0'1
`'""" Ul
`'""" ~
`~ c
`N c c
`
`'JJ.
`Cj
`
`00
`0 ......,
`N
`~ ......
`'JJ. =(cid:173)~
`
`~
`
`N c c
`
`~
`
`~N
`N
`~
`
`.... 0 =
`0' -....
`~
`.... 0 =
`~ 't:l -....
`~ = ......
`~ ......
`""C
`
`~ ......
`
`(')
`
`~ ......
`
`(')
`
`0.935
`
`0.126
`
`5.405
`
`1/llw/
`
`3.846
`
`0.788
`
`0.209
`
`1/llw/
`
`0.462
`
`1.290
`
`1.063
`
`1/llw/
`
`FINAL EQUIVALENT CHANNEL GAIN
`llwl
`X
`
`0.185
`
`X
`
`FINAL EQUIVALENT CHANNEL GAIN
`llw/
`0.260
`
`X
`
`X
`
`FINAL EQUIVALENT CHANNEL GAIN
`llwl
`X
`
`0.775
`
`X
`
`THIRD LAYER NULLING
`
`SECOND LAYER NULLING
`
`SECOND LAYER NULLING
`
`1.069
`
`X
`
`0.765
`
`llwl
`
`1
`
`o.425
`
`1.269
`
`X
`
`llw/
`
`X
`
`3.925
`
`0.941
`
`llwl
`
`SECOND LAYER NULLING
`
`FIRST LAYER NULLING
`
`THIRD LAYER NULLING
`
`7.941
`2.163
`LAYER
`LAYER
`SECOND THIRD
`
`4.789
`LAYER
`FIRST
`
`II wkll2
`
`2.163
`LAYER
`THIRD
`
`7.941
`LAYER
`SECOND
`
`4.789
`LAYER
`FIRST
`
`llw/
`
`7.941
`2.163
`LAYER
`LAYER
`SECOND THIRD
`
`4.789
`LAYER
`FIRST
`
`11 wkll2
`
`REVERSE ORDER FOR V-BLAST
`
`RANDOM ORDER
`
`ORDER FOR V-BLAST
`
`!
`
`1.3 0.2 1.2
`[
`H = 1.5 1.0 0.4
`1.2 0.5 1.5]
`
`FIG. 2
`
`SONY EX. 1005
`Page 3
`
`

`

`Patent Application Publication Jul. 22, 2004 Sheet 3 of 8
`
`US 2004/0141566 A1
`
`FIG. 3
`
`m = 1
`S = { 1,2, ... ,M}
`
`CALCULATES INVERTIBLE
`MATRIX OF H
`H~
`
`2 OF ALL K
`CALCULATES II wk 11
`( k E S)
`
`DECIDING LAYERS FOR NULLING
`k = argm x II wkll 2
`
`CORRECTING
`CHANNEL MATRIX AND SETS
`(H) k =>0,
`S => S-{k}
`
`31
`
`32
`
`33
`
`34
`
`35
`
`37
`m = m + 1
`
`DECIDING EQUIVALENT CHANNEL GAIN
`1/ 11'11!/. k = 1, ... ,M
`
`38
`
`SONY EX. 1005
`Page 4
`
`

`

`Patent Application Publication Jul. 22, 2004 Sheet 4 of 8
`
`US 2004/0141566 A1
`
`41
`
`42
`
`43
`
`44
`
`FIG. 4
`
`START
`
`" k = argmin Pk
`
`k
`
`C· = C· + 68
`k
`k
`
`END
`
`SONY EX. 1005
`Page 5
`
`

`

`'"""'
`>
`0'1
`0'1
`'"""' Ul
`'"""' ~
`~ c
`N c c
`
`'JJ.
`Cj
`
`00
`0 ......,
`Ul
`~ .....
`'JJ. =(cid:173)~
`
`~
`
`N c c
`
`~
`
`~N
`N
`~
`
`TRANSMITTING MIMO CHANNEL INFORMATION
`
`BIT AND CURRENT
`
`CALCULATOR
`INFORMATION
`ALLOCATION
`
`ESTIMATOR
`CHANNEL
`
`MIMO
`
`59
`
`58
`
`BIT AND CURRENT
`
`CALCULATOR
`INFORMATION
`ALLOCATION
`
`54
`
`DEMODULATING
`
`DATA FOR
`
`60
`
`DEMODULATION
`
`ADAPTIVE
`
`UNIT
`
`53-M
`
`MODULATION
`ADAPTIVE
`
`UNIT
`
`TRANSMITTING
`
`DATA FOR
`
`51
`
`.... 0 =
`0' -....
`~
`.... 0 =
`~ 't:l -....
`~ = .....
`~ .....
`""C
`
`~ .....
`
`(')
`
`~ .....
`
`(')
`
`57
`
`52
`
`56-1
`
`FIG. 5
`
`SONY EX. 1005
`Page 6
`
`

`

`Patent Application Publication Jul. 22, 2004 Sheet 6 of 8
`
`US 2004/0141566 A1
`
`FIG. 6
`
`-e- : SVD
`~ :BLAST
`-a- :BLAST
`__,._ :BLAST
`
`10-2
`
`10-3
`
`a:
`w
`co
`
`C]) co a: g
`
`w
`~ as
`w
`1-<r: a:
`a:
`0 a:
`a:
`w
`1-
`co
`
`104 ~--~----~----~----~--~n---~~--~--~
`-1
`0
`2
`3
`4
`5
`6
`7
`
`SIGNAL-TO-NOISE RATIO (Eb/N 0 ) (dB)
`
`SONY EX. 1005
`Page 7
`
`

`

`""""
`>
`0'1
`0'1
`"""" Ul
`"""" ~
`~ c
`N c c
`
`'JJ.
`Cj
`
`00
`0 ......,
`-..J
`~ .....
`'JJ. =(cid:173)~
`
`~
`
`N c c
`
`~
`
`~N
`N
`~
`
`.... 0 =
`0' -....
`~
`.... 0 =
`~ 't:l -....
`~ = .....
`~ .....
`""C
`
`~ .....
`
`(')
`
`~ .....
`
`(')
`
`CALCULATOR
`INFORMATION
`ALLOCATION
`
`BIT AND CURRENT
`83
`
`82
`
`ESTIMATOR
`CHANNEL
`
`MIMO
`
`DEMODULATED
`
`DATA
`
`84
`
`UNIT
`
`DEMODULATION
`
`ADAPTIVE
`
`81
`
`FFT
`I
`80-M
`
`80-2
`
`80-1
`
`TRANSMITTING MIMO CHANNEL INFORMATION
`
`BIT AND CURRENT
`73
`
`CALCULATOR
`INFORMATION
`ALLOCATION
`
`.-I -'~-o -,I
`
`IFFT
`I
`74-M
`
`MODULATION
`ADAPTIVE
`
`UNIT
`
`TRANSMITTING
`
`DATA FOR
`
`71
`
`72
`
`FIG. 7
`
`SONY EX. 1005
`Page 8
`
`

`

`Patent Application Publication Jul. 22, 2004 Sheet 8 of 8
`
`US 2004/0141566 A1
`
`FIG. 8
`
`-e-- : svo
`~ :BLAST
`--B- :BLAST
`~ :BLAST
`
`1 o-2
`
`10-3
`
`C2
`L1J m
`a.> co
`2 w
`~ co
`
`0::
`
`L1J
`1-
`<(
`0::
`0::
`0
`0::
`0::
`L1J
`1-
`m
`
`104 ~--~----~----~----~--~~--~--~~~--~
`-1
`0
`2
`4
`6
`5
`3
`7
`
`SONY EX. 1005
`Page 9
`
`

`

`US 2004/0141566 A1
`
`Jul. 22, 2004
`
`1
`
`APPARATUS AND METHOD FOR ADAPTIVELY
`MODULATING SIGNAL BY USING LAYERED
`TIME-SPACE DETECTOR USED IN MIMO
`SYSTEM
`
`FIELD OF THE INVENTION
`
`[0001] The present invention relates to an apparatus for
`modulating/demodulating signal by using a layered time(cid:173)
`space architecture detector in a multi-input multi-output
`(MIMO) system and a method thereof; and, more particu(cid:173)
`larly, to the apparatus for modulating/demodulating signal
`by using a layered time-space architecture detector in a
`multi-input multi-output (MIMO) system in order
`to
`increase a system performance by deciding an equivalent
`channel gain, deciding the number of bits for transmitting
`through each antenna and deciding a transmission power by
`using the a greedy algorithm based on the decided equiva(cid:173)
`lent channel gain.
`
`DESCRIPTION OF RELATED ARTS
`
`[0002] Generally, a multi-input multi-output (MIMO) sys(cid:173)
`tem is a wireless communication system obtaining high
`frequency efficiency by transmitting each different data
`through a plurality of transmission antennas in an identical
`bandwidth. There have been several detection methods for
`MIMO systems, such as a Diagonal Bell Laboratories Space
`Time (D-BLAST) system proposed in an article by G. 1.
`Foschini, entitled "Layered Space-Time Architecture for
`Wireless Communication in a Fading Environment When
`using Multi-Element Antennas,"Bell Labs Technical Jour(cid:173)
`nal, fall, pp. 41-59, 1999 and a Vertical-Bell Laboratories
`Space Time (V-BLAST) system introduced in an article by
`P. W. Wolniansky et.al., entitled "V-Blast: An Architecture
`for Realizing Very High Data Rates Over the Rich-Scatter(cid:173)
`ing Wireless Channel."Proc. International Symposium Sig(cid:173)
`nals, Systems and Electronics, September, 1998.
`
`[0003] The V-BLAST system is a modified scheme of the
`D-BLAST system.
`
`[0004] Operations of detection in the V-BLAST system
`are explained in more detail hereinafter. At first, a symbol
`corresponding to a layer having the largest equivalent chan(cid:173)
`nel gain is detected based on a MIMO channel matrix and
`new channel matrix is built by nulling a channel correspond(cid:173)
`ing to the currently detected layer after canceling the effect
`of the detected symbol in the channel matrix H. And the
`above mentioned steps are repeated until all symbols are
`detected. As a result of these operations, a layer having
`largest equivalent channel gain is detected for the first time.
`When the next symbol is detected, the effect of the first
`symbol is eliminated because it is considered as interference
`and then the next symbol is detected. Therefore, a diversity
`gain can be obtained when the next symbol is detected and
`a performance is increased.
`
`In a meantime, a structure of V-BLAST system is
`[0005]
`in U.S. Pat. No. 6,317,466 B1 issued to G. 1. Foschini,
`entitled "Wireless Communications System Having a Space(cid:173)
`Time Architecture Employing Multi-Element antenna at
`both the Transmitter and Receiver" and also an adaptive
`modulation method is disclosed. But in the above mentioned
`patent by G. 1. Foschini, a method for minimizing transmis(cid:173)
`sion power and changing modulation method commonly
`used for all antennas.
`
`[0006] Furthermore, a greedy algorithm for allocating
`power and the number of bits in multicarrier systems is
`introduced in an article by C. Y. Wong, entitled "Multi-user
`OFDM with adaptive sub-carrier, bit and power allocation",
`IEEE Journal on Selected Areas in Communication, Vol. 17,
`pp. 1747-1758, October, 1999. The greedy algorithm is used
`for calculating a power and the number of bits of each
`subcarrier in order to transmit all information bits with the
`minimum transmission power while satisfying desired bit
`error rate under conditions such as single user orthogonal
`frequency division multiplexing (OFDM) is used, a fre(cid:173)
`quency domain channel response corresponding to each
`subcarrier is known at the transmitter, the number of bits for
`transmitting at one OFDM symbol is predetermined and
`there is a desired bit error rate.
`
`[0007] There are two methods implementing the MIMO(cid:173)
`OFDM system with the adaptive modulation method. A first
`method is introduced in an article by Ka-Wai Ng. et. al.,
`entitled "A simplified bit allocation for V-BLAST based
`OFDM MIMO system in frequency selective fading chan(cid:173)
`nels"/£££ international conference on communication, pp.
`411-415, 2002. A second method is disclosed in an article
`by Ka-wai Ng et. al., entitled "Iterative bit & power allo(cid:173)
`cation for V-BLAST based OFDM MIMO system in fre(cid:173)
`quency selective fading channel", Proc. Wireless Commu(cid:173)
`nications and Networking Conference, 2002, pp. 271-275.
`
`[0008] The first method calculates the equivalent channel
`gain based on an ordering method of the V-BLAST and
`decides the number of bits according to the greedy algorithm
`based on the calculated equivalent channel. It is also dis(cid:173)
`closed that a method of selecting a subcarrier and transmit(cid:173)
`ting antenna in order to reduce the amount of information
`transmitted
`to
`the
`transmitter. The above mentioned
`approaches using the V-BLAST ordering method may obtain
`nearly same performance for all layers when fixed modula(cid:173)
`tion method is used. However, the performance is degraded
`in a case of an adaptive modulation is used.
`
`[0009] The second method allocates a predetermined
`number of bits to all possible combinations of decision order
`for the optimal solution and selects the optimal order with
`minimum total transmission power. However, by reducing
`the number of combinations to be considered for finding the
`optimal solution, the performance is degraded.
`
`SUMMARY OF THE INVENTION
`
`It is, therefore, an object of the present invention to
`[0010]
`provide an apparatus for adaptively modulating/demodulat(cid:173)
`ing signal by using a layered time-space architecture detec(cid:173)
`tor in a multi-input multi-output (MIMO) system for
`improving a system performance by deciding an equivalent
`channel gain, deciding the number of bits for transmitting
`through each antenna and a transmission power by using a
`greedy algorithm based on the decided equivalent channel
`gain.
`
`In accordance with an aspect of the present inven(cid:173)
`[0011]
`tion, there is provided an apparatus for adaptively modulat(cid:173)
`ing signal in a MIMO system having a layered space-time
`architecture based detector, the apparatus including: a bit
`and power allocation information calculator for deciding an
`equivalent channel gain-in a reverse order of Vertical-Bell
`laboratories Space Time (V-BLAS1) based on MIMO chan(cid:173)
`nel information feedbacked from a receiver and determining
`
`SONY EX. 1005
`Page 10
`
`

`

`US 2004/0141566 Al
`
`Jul. 22, 2004
`
`2
`
`the number of bits and transmission power to be transmitted
`from each transmitting antenna by using the equivalent
`channel gain; and adaptive modulation means for modulat(cid:173)
`ing signal of each layer with corresponding modulation
`method based on the determined number of bits and trans(cid:173)
`mitting power.
`[0012]
`In accordance with an aspect of the present inven(cid:173)
`tion, there is provided an apparatus for adaptively demodu(cid:173)
`lating signal in a MIMO system having a layered space-time
`architecture based detector, the apparatus including: MIMO
`channel estimation unit for estimating MIMO channel from
`a signal received through each receiving antenna; a bit and
`power allocation information calculator for determining an
`equivalent channel gain in reverse order of Vertical-Bell
`laboratories Space Time V-BLASTbased on MIMO channel
`information from the MIMO channel estimation unit and
`determining the number of bits corresponding to each trans(cid:173)
`mitting antenna by using the equivalent channel gain.
`[0013]
`In accordance with an aspect of the present inven(cid:173)
`tion, there is provided apparatus for adaptively modulating
`and demodulating signals in MIMO system using multiple
`antennas at transmitter and receiver, the apparatus including:
`an adaptive modulation unit for adaptively modulating sig(cid:173)
`nals in order to transmit the modulated signal after deter(cid:173)
`mining an equivalent channel gain in a reverse order of a
`vertical-bell laboratories space time and determining the
`number of bits and transmitting power based on the deter(cid:173)
`mined equivalent channel gain; and adaptive demodulation
`unit for detecting and adaptively demodulating received
`signals through each receiving antenna in reverse order of
`V-B LAST.
`[0014]
`In accordance with an aspect of the present inven(cid:173)
`tion, there is provided a method for adaptively modulating
`signals in a MIMO system using multiple antennas in a
`receiver and transmitter, the method, including the steps of:
`a) determining equivalent channel gain in a reverse order of
`V-BLAST at transmitter based on channel information feed
`backed from the receivers; and b) adaptively modulating
`signals by determining the number of bits and corresponding
`transmitting power to be transmitted through each layer
`(antenna) by using the equivalent channel gain in a greedy
`algorithm.
`[0015]
`In accordance with an aspect of the present inven(cid:173)
`tion, there is provided a method for adaptively demodulating
`in MIMO systems, the method including the steps of: a)
`estimating a channel from a signal received at each receiving
`antenna; b) deciding an equivalent channel gain in a reverse
`order of a vertical-bell laboratories space time (V-BLAST)
`based on the channel information; and c) detecting and
`adaptively demodulating the received signal using the adap(cid:173)
`tive modulation information for each layer.
`[0016]
`In accordance with an aspect of the present inven(cid:173)
`tion, there is provided a computer readable recoding
`medium storing instructions for executing a method for an
`adaptive modulation, the method including the steps of: a) at
`a transmitter, deciding an equivalent channel gain in a
`reverse order of V-BLAST based on a feedback information
`from a receiver; and b) at the transmitter, deciding the
`number of bit transmitting through each layer (transmitting
`antenna) and transmitting power based on the equivalent
`channel gain computed at the step a).
`[0017]
`In accordance with an aspect of the present inven(cid:173)
`tion, there is provided a computer readable recoding
`
`medium storing instructions for executing a method for an
`adaptive demodulation, the method including the steps of: a)
`estimating a channel from a signal received from each
`receiving antenna; b) deciding an equivalent channel gain in
`a reverse order of V-BLAST based on the channel informa(cid:173)
`tion; and c) detecting and adaptively demodulating by
`deciding the number bits for each layer.
`
`BRIEF DESCRIPTION OF 1HE DRAWING(S)
`[0018] The above and other objects and features of the
`present invention will become apparent from the following
`description of the preferred embodiments given in conjunc(cid:173)
`tion with the accompanying drawings, in which:
`[0019] FIG. 1 is a diagram for illustrating a MIMO
`wireless communication system having a layered space-time
`architecture detector in accordance with a preferred embodi(cid:173)
`ment of the present invention;
`[0020] FIG. 2 is a view showing changing of equivalent
`channel gain according to the detection order in layered
`space-time architecture in accordance with the present
`invention;
`[0021] FIG. 3 is a flowchart for explaining a method for
`determining the equivalent channel gain in a reverse order of
`the V-BLAST detection method;
`[0022] FIG. 4 is a flowchart for explaining steps for
`determining the number of bits for transmitting of each
`transmitting antenna and corresponding transmitting power
`by using a equivalent channel gain according to greedy
`algorithm in accordance with a preferred embodiment;
`[0023] FIG. 5 is a diagram showing an apparatus for
`adaptively modulating/demodulating signal in the MIMO
`wireless communication system having layered space-time
`architecture in accordance with a preferred embodiment of
`the present invention;
`[0024] FIG. 6 is a graph showing a performance of a
`method for adaptively modulating/demodulating signals in a
`MIMO wireless communication system having V-BLAST
`type detector in accordance with a preferred embodiment of
`the present invention;
`[0025] FIG. 7 is a diagram showing an apparatus for
`adaptively modulating/demodulating signals
`in MIMO
`OFDM wireless communication system having V-BLAST
`type detector in accordance with a preferred embodiment of
`the present invention; and
`[0026] FIG. 8 is a graph showing a result of a simulation
`of a method for adaptively modulating/demodulating signal
`in MIMO OFDM wireless system having V-BLAST type
`detector in accordance with a preferred embodiment of the
`present invention.
`
`DETAILED DESCRIPTION OF THE
`INVENTION
`[0027] Other objects and aspects of the invention will
`become apparent from the following description of the
`embodiments with reference to the accompanying drawings,
`which is set forth hereinafter.
`[0028] For helping to understand the present invention, a
`greedy algorithm used in an adaptive modulation/demodu(cid:173)
`lation method in accordance with the present invention is
`explained at first.
`
`SONY EX. 1005
`Page 11
`
`

`

`US 2004/0141566 Al
`
`Jul. 22, 2004
`
`3
`
`[0038]
`
`is satified.
`
`[0029] For explaining the greedy algorithm, an orthogonal
`frequency division multiplexing method is used as an
`example.
`
`[0030] At first, the transmission power of the n'h subcarrier
`of, P n• is given as:
`
`!:J.P = :_f_( C-"-n _+--cllc-Bcc) ----'--! (_C_c_n)
`lhnl 2
`n
`
`Eq. 1
`
`Eq. 3
`
`Eq. 4
`
`Eq. 5
`
`In the Eq. 1, hn is a channel gain of the n'h
`[0031]
`subcarrier, en is the number of bit for transmitting with the
`subcarrier, and f(en) is a is a transmission power for
`transmitting c bits with desired bit error rate (BER).
`
`In a meantime, if an entire bits for transmitting one
`[0032]
`of OFDM symbol is R, a bit allocation method for mini(cid:173)
`mizing a sum of transmission power of entire subcarrier can
`be expressed as:
`
`Eq. 2
`
`[0033]
`
`subject to
`
`In the Eq. 2, Nc is the number of total subcarriers,
`[0034]
`D represents a set of information bits per subcarrier deter(cid:173)
`mined by modulation/demodulation methods. For example,
`if a QPSK, 16-QAM, 64-QAM and so on can be selected for
`the modulation/demodulation methods, then D={O, 2, 4, 6,
`8, ... } . In here, 0 means that no information is sent in the
`corresponding subcarrier.
`
`[0035] The Eq. 2 can be solved by the greedy algorithm
`through following steps.
`
`[0036]
`
`step 1: For all subcarriers, we define cn=O,
`
`f(llB)
`!:J.Pn = - -2 .
`lhnl
`
`step 2: following equations 3, 4 and 5 are
`[0037]
`repeatedly operated until the condition
`
`In the Eqs. 4 and 5, llB is the difference of elements
`[0039]
`of set D e.g., D={O, 2, 4, 6, ... }. Therefore, llB=2.
`
`step 3: after completing step 2, each en is
`[0040]
`determined and P n is calculated by using eq. 1.
`
`[0041] Finally, the number of bits allocated to each sub(cid:173)
`carrier, en, and the transmission power, p n• is determined.
`
`In the V-BLAST detection method, data transmit(cid:173)
`[0042]
`ted from a plurality of transmitting antenna is detected as
`one by one and the detected symbols are cancelled for
`reducing interference to other signals. The next symbol is
`detected after canceling the channel effect corresponding to
`the previously detected symbol. Therefore, the equivalent
`channel gain is varied according to an order of nulling. The
`present invention provides a method for achieving perfor(cid:173)
`mance gain by adaptively modulating the signal with an
`equivalent channel gain obtained by the reverse order of
`conventional V-BLAST detection order. Hereinafter, the
`V-BLAST detection method is explained at first.
`
`[0043] FIG. 1 is diagram for illustrating a MIMQ wireless
`communication system having a layered space-time archi(cid:173)
`tecture detector in accordance with a preferred embodiment
`of the present invention.
`
`[0044] Referring to FIG. 1, in the MIMO wireless com(cid:173)
`munication system having a detector of V-BLAST structure,
`a signal vector transmitted from each transmission antenna
`13-1, 13-2, ... , 13-M is defined as x=[x0 , x1 , . . . , xM_ 1y
`and a signal vector received to each receiving antenna 15-1,
`15-2, ... 15-N is defined as y=[y0 , y1 , . . . , YN-lY· M and
`N is the number of transmitting/receiving antennas and it is
`satisfied with a condition, N~M. The received signal vector
`is expressed as:
`
`Eq. 6
`y=Hx+v
`[0045]
`In the Eq. 6, His NxM channel matrix, the element
`ofthe n'h row and, a m'h column, hn m• is a fiat fading channel
`gain between m'h transmitting a~tenna and n'h receiving
`antenna. In a meantime, vis a Nx1 white noise vector with
`zero mean and its covariance matrix is E[ vvH]=a2I . I is
`NxN unit matrix.
`
`[0046] For detection of the transmitting signal vector x
`from the receiving signal vector y, the V-BLAST detection
`method is used. A nulling vector for k'h layer (k'h transmis(cid:173)
`sion antenna signal) is defined as:
`
`w[(H)1 = bkl =
`
`1
`0
`
`{
`
`, k = l
`
`, k * l
`
`Eq. 7
`
`In the Eq. 7, (H)l is l'h column of matrix Hand wk T
`[0047]
`is a k'h column of a pseudo-inverse matrix H+. A detection
`
`SONY EX. 1005
`Page 12
`
`

`

`US 2004/0141566 Al
`
`Jul. 22, 2004
`
`4
`
`order is determined by considering a value of wk. In the
`conventional V-BLAST detection, a column of the matrix
`having smallest value of llwkll is detected at first. In a
`meantime, ~ is a decision statistic of k'h layer and it is
`expressed as:
`
`Eq. 8
`zk~xk+wkTV
`[0048]
`In the Eq. 8, k={ 1, 2, ... , M}. After determining
`the signal by considering the decision statistic ~. an inter(cid:173)
`ference of the previous detected signal is cancelled and the
`channel matrix H is updated. That is, new channel matrix H
`is generated by setting k'h column of prior matrix H as '0'.
`New received signal vector y' is determined as:
`
`~9
`y~~~
`[0049]
`In the Eq. 9, xk is a result of decision of~· The
`Eqs. 6 to 9 are repeatedly executed for calculating a nulling
`vector by using new channel matrix H' and receiving signal
`vector y' until all signals are detected. The above mentioned
`steps are detection steps of conventional V-BLAST. Order of
`detection is determined according to the magnitude of norm
`of nulling vector.
`
`If the eq. 7 is used for calculating a signal-to-noise
`[0050]
`ratio of the decision statistics, Pk> which is calculated as:
`
`E{lxdJ
`Pk ~ cr2llw, 11 2
`
`Eq. 10
`
`In the, Eq. 10, E{} is an expectation function and
`[0051]
`llwkll 2 is the equivalent channel gain. Therefore, a state of
`channel is better as llwkll is getting smaller. The V-BLAST
`method for MIMO system using a fixed modulation/de(cid:173)
`modulation method detects a layer of best channel state at
`first and then detects a layer of worse channel state after
`eliminating interference of signals, which have already been
`detected in order to enhance performance of all layers.
`
`[0052] However, the adaptive modulation/demodulation
`method detects a layer of best channel state later and
`information of a layer of worse channel is not transmitted or
`less amount of information is transmitted in order to enhance
`the performance. Therefore, the detection order must to be
`reversed for enhancing the performance in accordance with
`the present invention.
`
`[0053] FIG. 5 is a diagram showing an adaptive modula(cid:173)
`tion/demodulation, and in the MIMO wireless communica(cid:173)
`tion system with V-BLAST type in accordance with a
`preferred embodiment of the present invention.
`
`[0054] Referring to FIG. 5, a MIMO system having
`V-BLAST detector includes an adaptive modulation unit 52
`and an adaptive demodulation unit 57 for modulating/
`demodulating signals with using different demodulation/
`modulation method and controlling transmission power of
`the signal. The demodulation unit 57 includes a V-BLAST
`type detection unit.
`
`[0055] Operations of the adaptive modulation/demodula(cid:173)
`tion units 52 and 57 are explained in detail as followings.
`
`ting power by modulating the data with different modulation
`methods according to each layer by using bit and power
`allocation information.
`
`[0057] When adaptively modulated signals are transmitted
`through M transmitting antennas 53-1 to 53-M, a bit and
`power allocation information calculator 54 determines the
`number of bits and transmitting power to be transmitted to
`each transmitting antennas 53-1 to 53-M by using MIMO
`channel information feedbacked from a receivers and the
`greedy algorithm.
`
`[0058] Through the above mentioned operations, signals
`transmitted from theM transmitting antennas 53-1 to 53-M
`are transmitted to N receiving antennas 56-1 to 56-N
`through frequency non-selective fading MIMO channel.
`And the adaptive demodulation unit 57 demodulates the
`received signals by using channel estimation result from a
`MIMO channel estimator 58 and a bit allocate information
`from the bit allocation information calculator 59 in reverse
`order of V-BLAST. Finally, the demodulated signal is out(cid:173)
`putted.
`
`[0059] FIG. 2 is a view showing changing of equivalent
`channel gain according to an order of nulling in a layered
`space-time architecture in accordance with the present
`invention.
`
`[0060] Referring to FIG. 2, although identical channel
`matrix is used, the equivalent channel gain is changed
`according to an order of detection such as an order of
`V-BLAST, a random order and a reverse order ofV-BLAST.
`Thus, in case of detection according to the order of
`V-BLAST, a deviation of the equivalent channel gain is
`small and the sum of equivalent channel gain is small. And
`in case of detecting according to the reverse order of
`V-BLAST, the deviation and sum of the equivalent channel
`gain become larger.
`
`[0061] A method for determining the equivalent channel
`gain in a reverse order of the V-BLAST detection method in
`accordance with a preferred embodiment of the present
`invention is explained in detail.
`
`[0062] FIG. 3 is a flowchart for explaining the method for
`determining the equivalent channel gain in a reverse order of
`the V-BLAST detection method.
`
`[0063] Referring to FIG. 3, at step 31, an antenna index
`set is defined (S={ 1, 2, ... , M}) because of determining an
`order of detection for determining the equivalent channel
`gain. An pseudo-inverse Matrix of channel matrix H+ is
`calculated at step 32. At step 33, all nulling vectors wk T, kES,
`are calculated based on the Eq. 7 for calculating transmitting
`signal vector x.
`
`[0064] After the step 33, square root of a norm of nulling
`vector is calculated for determining a layer for nulling and
`k'h layer having the largest value is selected at step 34. A
`column of the channel matrix corresponding to the selected
`k'h layer is transformed to 0 vector and a value k is
`eliminated from the set S at step 35. The above mentioned
`steps are repeated M times with newly transformed setS and
`channel matrix H.
`
`[0056] A transmitting data 51 is data transmitted from a
`transmitter. If necessary, this date would be scrambled,
`channel coded, and interleaved. The adaptive modulation
`unit receives the channel information and controls transmit-
`
`[0065] As mentioned above, an order of the selected layers
`becomes a detection order and an inverse of square root
`norm of the nulling vector for the selected layer becomes the
`equivalent channel gain. Furthermore, the number of bits
`
`SONY EX. 1005
`Page 13
`
`

`

`US 2004/0141566 Al
`
`Jul. 22, 2004
`
`5
`
`transmitted at each layer (transmitting antenna) and corre(cid:173)
`sponding transmitting power is calculated by using the
`calculated equivalent channel gain instead of subcarrier
`channel gain llhnll 12 in greedy algorithm.
`[0066] FIG. 4 is a flowchart for explaining steps for
`determining the number of bits for transmitting of each
`transmitting antenna and corresponding transmitting power
`by using a equivalent channel gain according to greedy
`algorithm in accordance with a preferred embodiment of the
`present invention.
`[0067] Referring to FIG. 4, at step 41, the number of bits
`transmitted at each layer (each transmitting antenna) is
`initialized as '0'. llB is defined based on the Eq. 2 and a
`transmitting power is calculated for obtaining an additional
`bit value of llB with desired bit error rate at each layer by
`using the equivalent channel gain of each layer. Among the
`calculated layers, a layer requiring the least transmitting
`power is selected and the value of llB is putted to the
`selected layer. The above mentioned steps are repeatedly
`performed until allocated total bit number is assigned. After
`determining the number of bits transmitted through each
`layer, a transmitting power of each layer is calculated based
`on the Eq. 1.
`[0068] FIG. 6 is a graph showing a performance of an
`adaptive modulation/demodulation method of a MIMO
`wireless communication system with V-BLAST type detec(cid:173)
`tor in accordance with a preferred embodiment of the
`present invention. The graph shows result of simulation for
`comparing performance of a conventional modulation/de(cid:173)
`modulation method and
`the modulation/demodulation
`method in accordance with the present invention.
`
`In the simulation, the number of transmitting and
`[0069]
`receiving antennas is 4 and a channel of each transmitting
`and receiving antenna is a complex Gaussian random vari(cid:173)
`able with zero mean. For preventing to limit performance by
`influence of specific channel matrix, more than 1000 statistic
`channels are generated. The number of information bits for
`transmitting in a predetermined time is set to 8 and QPSK
`and 16-QAM are used for modulation method. Thus, D={O,
`2, 4} and therefore, llB is 2.
`
`In the simulation, a system performance of a con(cid:173)
`[0070]
`ventional modulation/demodulation method implemented
`by using greed algorithm with singular